We all saw it. None of us will be able to forget it, unless, of course, Houston wins tonight. With a four-run lead and needed just six outs to clinch an opportunity to play for a pennant, the Astros bullpen utterly blew up, allowing seven runs and sending the series back to Kansas City for a do-or-die Game 5 showdown.
During the debacle, the FS1 broadcast crew discussed to some length the lack of hard-throwing relievers in the Astros bullpen. I'll be frank with you upfront; by that time, I was literally so numb that I wasn't able to pay attention very closely, but the overall jist of the discussion was that Houston's lack of bullpen velocity had them at a disadvantage, and maybe this wouldn't be happening if they had more guys who could throw hard.
Following the game, Evan Drellich over at the Chronicle penned this piece which follows the same narrative. His overarching point appears to focus on the idea that Houston lacks both velocity out of the pen, and that, because of that, at least in part, they don't strike batters out. One line in particular boils down the aim of the entire article pretty succinctly:
Astros relievers don't miss bats. Royals hitters don't miss balls. It's a bad combination for the former.
Drellich brings up Andrew Miller spurning the club's offer in the off-season, and also their failed attempts to trade for one of either Aroldis Chapman or Craig Kimbrel at the trade deadline. It's clear that the Astros hunted for some fire to add to the bullpen and came up short.
The conclusion being drawn, however, seems a bit broad. To his credit, Drellich goes on to discuss how it was a lack of execution, at heart, that cost the Astros bullpen so dearly. He simply makes the case that relievers who throw harder can get away with more mistakes on account of their velocity being harder to hit.
So I did some digging. A few notes; first, that "fastballs" include four-seamers, two-seamers, sinkers, and cutters, unless I specify otherwise; and second, I decided to use xFIP as my base on how well a pitcher performed; aside from xFIP generally being a better stat than ERA, I believe it is especially for single-season reliever stats, as those are inherently smaller sample sizes.
Sorting for the top 50 best (lowest) xFIP totals among qualified relievers, five of the top 10 (50%), 13 of the top 25 (52%) and 29 of the top 50 (58%) have a fastball that averages 94 MPH or higher. Interesting already is that the percentage of big fastballs seems to increase as you get towards worse (higher) xFIPs, but it's not a huge difference and the sample isn't big, so maybe it's just noise.
Let's flip the table; one of the bottom ten (10%), four of the bottom 25 (16%) and 11 of the bottom 50 (22%) relievers in terms of xFIP had the 94 MPH average for their fastballs. Pretty big difference that would seem to indicate that having a big heater helps.
So how about the guys in the middle? Well, there were 137 total qualified relievers, so that would be 51 through 87 on the list. 15 of those 37 guys (40.5%) had the big velocity in their tool box. Nine of the 14 (64.3%) guys just before you drop into the bottom 50 had the 94 MPH heat, though, a better percentage than even the top 10 elite relievers.
So what does this mean? Well, it would seem that the big velocity is nice. It certainly can help. But there's not a huge separation between the middle-road guys and the top guys, not even the elite. You could say that it helps you not to be bad, but these numbers don't strongly suggest that you need it to be elite, or that it's what separates the best from the average.
I didn't just to a straight-up comparison of the two, though. The article makes mention of strike outs and getting away with mistakes. The first is easy; we look at strike out rates. The second is a little harder. I'm going to look at Contact% and line drive rate (LD%) because, ostensibly, the bigger fastballs mean that guys are less likely to make contact and do damage with said contact when the pitchers make mistakes than they would if the pitcher didn't have the heat, right?
Sorting for the 50 best (lowest) Contact% relievers, six of the top 10 (60%), 12 of the top 25 (48%), and 24 of the top 50 (48%) had the 94 MPH cheese. Flipping the table again, one of the bottom 10 (10%), one of the bottom 25 (4%), and ten of the bottom 50 (20%) could consistently pump the gas. Note that that means just one of the 25 worst guys at allowing contact had the heat, but nine of the next 25 did.
And for the middle-road group; well, 20 of those 37 guys (54.1%) of them had big velocity. Five of the six just outside the bottom 50 had it. Clearly, then, it seems to hold true; as far as allowing contact, velocity can help you not be bad, but it doesn't necessarily mean you'll be elite, or even above-average.
Five of the top 10 (50%) worst at allowing line drives had the big fastball, as did eight of the bottom 25 (32%), and 18 of the bottom 50 (36%). Meanwhile it was three of the top 10 (30%), eight of the top 25 (32%), and 19 of the top 50 (38%) who could call the thunder. Far from allowing fewer line drives, it appears that big velocity might actually make you a little more susceptible to liners. Once again, little evidence here that velocity is the key to relief dominance.
There's a pretty solid correlation between strike out rate and xFIP; six of the 50 worst xFIP guys (12%) struck out 10 batters or more per nine innings, while 30 of the top 50 xFIP guys (60%) posted a 10.00 or better K/9. However, only six guys in that middle 37 (16.2%) managed the 10 strike outs per nine innings. To put it another way, the 60% of the top 50 xFIP guys had the big strike out rate, while just 13.8% of the remaining 87 relievers did. That's notable; it seems strike outs could indeed play a big role in reliever success after all.
However, there were 47 relievers who struck out fewer than eight batters per nine, and nine of them (19.2%) had a big heater. On the other side, of the 42 relievers who struck out 10 batters per nine innings or better this year, 25 (59.5%) had a big heater. The guys that whiffed between 9.99 and 8.00 batters per nine numbered 48, and 21 of them (43.8%) had the hard cheese in their lunch pail. This time there's a decent separation between the biggest strike out guys and the middle-road ones, but it's not nearly as big as the one between the bad ones and the average.
Again, velocity certainly helps, but it doesn't look to be the end-all, be-all. Given the gap in strike out rate between the best performing relievers, you'd expect a bigger correlation between velocity and strike outs if it was velocity that really lead to more strike outs. So the train of thought that velocity leads to strike outs leads to dominance doesn't seem to hold up when looking at the stats. Strike outs seem to lead to dominance, but there's not a great deal of evidence that velocity definitely begets more strike outs, or even just less contact. It seems there's a point of diminishing returns; velocity can keep you from being at the bottom of the barrel, but you'll likely need more than that if you want to be elite.
A few other tidbits in closing; while it's true that the Royals don't strike out (lowest team K% in the Major this year, in fact), the bit about the Astros relievers not missing bats simply isn't true; Astros relievers posted a combined 9.11 K/BB this year, the fifth-best mark in baseball. Their combined Contact% was the eleventh-best mark in the Majors, too, so the idea that they can't miss bats is just incorrect.
By the way, the Astros beat the Royals handily in combined reliever xFIP, strike out rate, walk rate, strikeout-to-walk ratio, and Contact%. In fact, they're more than eight places better than Kansas City in all of those categories but Contact%. Just something to remember for later tonight if you hear some broadcasters start swooning over the Royals bullpen while lamenting the lack of big velocity in the Astros'.